File: `README.txt`

Dataset Title: Dataset for the article: "OPEN-SOURCE AI REVIEWER: A UNIVERSAL PROMPT FOR HYBRID EXPERTISE AND SELF-ANALYSIS OF SCIENTIFIC TEXTS"  

Author: Kravtsov Gennady Grigorievich  
ORCID: https://orcid.org/0009-0000-3405-1461  
Related Publication: [https://doi.org/10.5281/zenodo.17020322]  
License: Creative Commons Attribution 4.0 International (CC BY 4.0)  

OVERVIEW  

This dataset contains supplementary materials for the article "OPEN-SOURCE AI REVIEWER...". The materials demonstrate the practical application of the proposed universal AI review prompt and serve to ensure transparency, verifiability, and reproducibility of the research.  

FILE DESCRIPTIONS  

1.  `README.txt` (this file)  
    - Content: Description of the dataset contents.  

2.  `Universal_AI_Review_Prompt_v_31_EN.txt`  
    - Content: Text file containing the open-source prompt.  
    - Description: The final, tested version (v.3.1) of the universal prompt. Designed for use with AI assistants (e.g., DeepSeek Chat) to conduct structured peer review of scientific texts.  

3.  `Kravtsov_AI_Reviewer_Open_Source_Article_EN.pdf`  
    - Content: Scientific article "AI REVIEWER WITH AN OPEN PROMPT: A UNIVERSAL PROMPT FOR HYBRID EXPERTISE AND SELF-ASSESSMENT OF SCIENTIFIC TEXTS".  
    - Description: The main theoretical article describing the methodology, development, and testing of the universal AI review prompt. Provides comprehensive background and research context.  

4.  `AI_Review_Open_Source_Article_EN.pdf`  
    - Content: AI-generated review using the prompt.  
    - Description: An example review automatically generated for the article "OPEN-SOURCE AI REVIEWER...". Serves as a demonstration of the prompt's functionality and a reference for future users.  

5.  `Comparative_Table_4_Articles_Quality.pdf`  
    - Content: PDF document with a comparative table.  
    - Description: A summary table automatically generated by the AI, providing a comparative analysis of the quality of four scientific articles based on standardized criteria. Clearly demonstrates the method's capability for comparative analysis.  

INSTRUCTIONS FOR USING THE PROMPT  

1. Copy the contents of the file `Universal_AI_Review_Prompt_v_31_EN.txt`.  
2. Open a new chat in a supported AI assistant (e.g., DeepSeek Chat).  
3. Paste the copied prompt text.  
4. Add "Scientific article for review" and insert the full text of the scientific article to be analyzed.  
5. Add necessary references to supplementary materials and their descriptions.  
6. Execute the prompt.  
7. For optimal results, it is recommended to run the process multiple times and consider the averaged outcome.  

KEYWORDS  

artificial intelligence, scientific peer review, open science, prompt engineering, AI assistant, manuscript self-analysis, hybrid expertise, machine text analysis, prompt engineering, AI assistant, peer review, source code  

CONTACT INFORMATION  

For any inquiries regarding this dataset, please contact:  
Gennady Kravtsov, Director of the Research Center "Applied Statistics".  
Email: 62abc@mail.ru